The Effects of Extracorporeal Cardiopulmonary Resuscitation According to Covariate Adjustment

Abstract

This study compared the effects of extracorporeal cardiopulmonary resuscitation (ECPR) using propensity-score matching (PSM) analyses. A nationwide registry of out-of-hospital cardiac arrest (OHCA) patients in Korea between 2013 and 2016 was used. Patients with OHCA aged ≥15 years with cardiac etiology and resuscitation time >30 minutes were enrolled. Resuscitation-related variables before the initiation of ECPR were included. Two PSM analyses were performed separately, with and without post-ECPR variables. The primary outcome (PO) was a favorable neurologic outcome at hospital discharge. The rate of PO was 8.1% (13/161) in the ECPR group and 1.5% (247/16,489) in the conventional CPR (CCPR) group. In the matched cohort with post-ECPR variables, there was no significant difference in the rate of PO between the ECPR and CCPR groups (7.9% vs. 7.9%; p = 0.982). In the matched cohort without post-ECPR variables, the rate of PO was higher in the ECPR group than that in the CCPR group (8.3% vs. 3.6%; p = 0.012). PSM analysis without post-ECPR variables compared outcomes of all patients experiencing OHCA and treated with ECPR versus CCPR, which showed better neurologic outcomes for ECPR. PSM analysis with post-ECPR variables compared outcomes between ECPR survivors and CCPR survivors, which exhibited similar neurologic outcomes.